Dismantling Lamarckism: Why Descriptions of Socio-Economic Evolution as Lamarckian are Misleading

Geoffrey M. Hodgson and Thorbjørn Knudsen

20 February 2006

Published in the Journal of Evolutionary Economics

The Business School, University of Hertfordshire, De Havilland Campus, Hatfield, Hertfordshire AL10 9AB, UK

http://www.herts.ac.uk/business http://www.geoffrey-hodgson.ws

and

Department of Marketing & Management, University of Southern Denmark, Odense Campus, DK-5230, Odense M,

Denmark

Address for correspondence:

Malting House, 1 Burton End, West Wickham, Cambridgeshire CB1 6SD, UK

g.m.hodgson@herts.ac.uk

KEY WORDS:

Lamarckism, Darwinism, evolution, genotype, phenotype, memes, habits, routines

JEL classification: B52, D20, D83

ABSTRACT

This paper addresses the widespread tendency to describe socio-economic evolution as

Lamarckian. The difference between Lamarckian and Darwinian replication is clarified. It is

shown that a phenotype-genotype distinction must first be established before we can identify

Lamarckian transmission. To qualify as Lamarckian inheritance, acquired properties at the

phenotypic level must be encoded in a genotype that is passed on to the next generation. Some

possible social replicators (or genotypes) are identified, with a view to exploring possible

distinctions between genotype and phenotype at the social level. It is concluded that the

Lamarckian label does not readily transfer to socio-economic evolution, despite the fact that

social genotypes (such as routines) may adapt within any given phenotype (such as an

organisation). By contrast, no such problems exist with the description of socio-economic

evolution as Darwinian.

- 1 -

Dismantling Lamarckism: Why Descriptions of Socio-Economic Evolution as Lamarckian are Misleading

Geoffrey M. Hodgson and Thorbjørn Knudsen

1. Introduction

Many prominent social scientists including Jack Hirshleifer (1977), Herbert Simon (1981),

William McKelvey (1982), Richard Nelson and Sidney Winter (1982), Robert Boyd and Peter

Richerson (1985), Friedrich Hayek (1988), and Arthur Robson (1995), have described socio-

economic evolution as „Lamarckian‟.1 Unfortunately, the precise meaning of the term, and

whether or not it is meant to exclude Darwinism, is less frequently made clear. Critics of this

Lamarckian label are few, including David Hull (1982, 1988) and John Wilkins (2001).

Wilkins (2001) portrays it as an ambiguous term with three prominent and different meanings:

The first meaning of Lamarckism is the notion that acquired characters can or will be

inherited. Jean Baptiste de Lamarck strongly promoted this idea, but it was not original to

him.2 We discuss this notion of Lamarckism extensively below.

A second strong theme in the writings of Lamarck, which he developed rather than

originated, is the idea that evolution involves increasing complexity. Although later

Lamarckians such as Herbert Spencer took up this idea, it has today grown beyond its

Lamarckian associations, and is rarely associated specifically with the Lamarckian label.

1 The authors wish to thank Marion Blute, Peter Corning, Kevin Greene, David Hull, John Nightingale, Richard

Nelson, the editor Uwe Cantner and two anonymous referees for comments on earlier versions. This paper is

dedicated to David Hull, who inspired much of its argument.

2 Lamarck ([1809] 1984, p. 113) himself believed in the stronger version of this thesis, that all acquired

characters are inherited: „All the acquisitions and losses wrought … through the influence of the environment …

are preserved by reproduction to the new individuals‟.

- 2 -

A third use of the Lamarckian label associates it with the emphasis on will, choice,

anticipation, or volitional activity in the process of evolutionary change. Many Lamarckians

have invoked the concept of will or volition to explain the development of acquired

characteristics (e.g. Butler, 1878). However, Lamarck himself emphasised neither will nor

volition, and their association with Lamarck‟s theory originates with his hostile critic Georges

Cuvier (Lamarck, [1809] 1984; Boesiger, 1974; Burkhardt, 1977).

In an obvious and general sense, the third meaning of Lamarckism is uncontroversial and

does not exclude Darwinism. Although human mental capacities are more highly developed,

most living organisms anticipate, choose, and strive for prefigured goals. These intentional

factors generally play a major role in biological as well as cultural evolution, because the

nature and sophistication of these cognitive mechanisms has an enormous bearing on

adaptation and survival in the evolutionary process. This was a theme in Darwin‟s own

writings, and has been developed by leading Darwinian biologists (Mayr, 1960; Waddington,

1969, 1976; Corning, 1983). Darwinism does not deny intentionality, it simply insists that it

has evolved in a causal process, and that intentions themselves are caused.

It is only in a special case, where these anticipative and purposive capacities are assumed to

have somehow appeared independently of a Darwinian evolutionary process, that the third

meaning becomes problematic. Yet this extraordinary version of the third meaning, with its

open door to theism, is as far from Lamarck as one could imagine.

In its uncontroversial form, the third meaning acquires more bite when it is combined with

the first meaning of Lamarckism above: volition thus becomes part of the mechanism by

which new characteristics are developed and acquired. But the third meaning says nothing

about inheritance, which is the key element in the first meaning.

The volitional acquisition of characteristics is often contrasted to allegedly „blind‟ or

random mutations in some versions of Darwinism. In response, Darwin himself never wrote

of random mutations, and in principle core Darwinian principles are broad enough to

accommodate both contrasting accounts. Furthermore, the acquisition of characters is

primarily a developmental phenomenon,3 whereas genetic mutations necessarily involve

changes in a genotype. For the comparison to be appropriate, it has at least to be upheld that

- 3 -

acquired characters are encoded in the genotype and that the modifications in the genotype

are passed on to offspring. In fact, Darwin (1859, 1868) himself believed in both these

possibilities.

We accept the possibility, in the social if not the biological sphere, that a (social) genotype

may be affected by a (social) phenotype, just as firms can affect their routines in the Winter

(1971) and Nelson and Winter (1982) models. Our argument below concerns problems in the

other part of the argument, concerning the inheritance of acquired characters. Hence the

controversy surrounding the Lamarckian label centres mostly on the first meaning.

In two independently drafted papers (Hodgson, 2001; Knudsen, 2001), we asked if socio-

economic evolution were Lamarckian or Darwinian? We upheld that both biological and

social evolution are Darwinian, in the sense that Darwin‟s principles of variation, selection,

and inheritance apply to both biological and socio-economic entities. This does not mean that

the details of the evolutionary processes are similar; it means simply that broadly defined and

generalized Darwinian principles apply to both domains (Hodgson, 2002; Hodgson and

Knudsen, 2006). The idea that general Darwinian principles apply to social as well as

biological evolution was suggested by Darwin (1859, 1871) himself and has been put forward

by a number of subsequent authors.4 The related insight that Darwinian principles apply to

any level of biological organization, including molecules, cells, organisms, groups, species,

and ecological communities was introduced by Lewontin (1970). Today, it is widely accepted

(Brandon, 1999).

We also pointed out that the proposition that socio-economic evolution is broadly

Darwinian does not mean that a Lamarckian possibility is necessarily excluded. However,

since the theoretical and experimental work of August Weismann (1893) this possibility has

3 The organism‟s acquisition of characters in developmental or mature stages is commonly referred to as

phenotypic plasticity.

4 Notably Campbell (1965) and Dawkins (1983). See Hodgson (2004) for a historical account.

- 4 -

been generally excluded in modern biology, because there is no apparent mechanism by

which the acquired characters of an Earthly organism can be transferred to its genotype.5

But the units and mechanisms of socio-economic evolution are very different in nature and

in human society. Does Lamarckian inheritance apply widely to the human but not the

biological world? There is nothing in the general Darwinian principles that logically rules out

this possibility; Lamarckism and Darwinism are not mutually exclusive. This is confirmed by

inspection of the following definitions of these terms:

Darwinism is a causal theory of evolution in complex or organic systems, involving

the inheritance of genotypic instructions by individual units, a variation of genotypes,

and a process of selection of the consequent phenotypes.6

Lamarckism is a doctrine admitting the possibility of the (genotypic) inheritance of

acquired (phenotypic) characters by individual organisms in evolutionary processes.

Weismannism (or neo-Darwinism) is a doctrine denying the possibility of the

(genotypic) inheritance of acquired (phenotypic) characters by individual organisms in

evolutionary processes.

In our previous papers on this topic, after elaborating some of these arguments in detail, we

concluded that the question of whether or not Lamarckian inheritance occurs in socio-

economic evolution would have in part to be answered by empirical enquiry into the actual

5 There is a minority view among biologists that the inheritance of acquired characters may be possible in a

restricted set of circumstances, such as the transfer of acquired immunities from mother to child (Steele, 1979;

Ho and Saunders, 1984; Jablonka et al., 1992; Steele et al., 1998). We entirely abstain from evaluating these

minority arguments in biology, and our argument here would be unaffected by either their validity or their

falsehood. Our imaginary journey (see below) to Planet Lamarck is a thought experiment, asking the question: if

Lamarckian inheritance existed, then what would be involved? It does not mean that we believe in the possibility

of Lamarckian inheritance in biological organisms on Earth. And concerning life on other real planets, we have

insufficient knowledge to form an opinion.

6 In biology, the genotype is the complete genetic coding of an organism, consisting of instructions on how it

should grow and develop. Many of these instructions depend upon environmental triggers or stimuli. The

phenotype is its actual character, including its behavioural propensities and capabilities. Each individual

phenotype develops according to the instructions in its genotype and the influence of environmental conditions.

As noted below, there is a close parallel here with the allied concepts of replicator and interactor.

- 5 -

mechanisms of replication involved. If it did occur, it was argued, it would have to ensure

some continuity and durability of the accumulated knowledge embodied in habits and

routines, and the Lamarckian process would necessarily rely on the complementary

Darwinian mechanism of selection. The overall outcome of our earlier discussions was to

leave the question of the extent of Lamarckism in socio-economic evolution open. In the

present article we go further, by considering the conceptual limits to Lamarckian socio-

economic evolution in more depth.

The popularity of the Lamarckian description of socio-economic evolution is extraordinary.

It is not the aim of the paper to explain this popularity: instead we concentrate on the

problematic nature of the label. To some it may appear obvious that key human phenomena

such as learning allow the development and transmission to subsequent generations of

adaptations much more rapidly than among other species. For some, this difference

substantiates the description of socio-economic evolution. We fully accept that cultural

transmission in human societies occurs and is much more important than in other species.

Human culture is unique in its nature, dimensions and significance.

However, such observations do not themselves justify the Lamarckian label. Lamarckism

involves the inheritance of acquired characteristics. Inheritance means more than merely

„passed on‟. If it were merely the latter then the spread of a virus among members of any

species would be evidence of Lamarckism. No biologist regards such epidemiological

contagions as Lamarckian. The concept of inheritance is and must be invested with a different

meaning: it must involve transmission from some kind of genotype to another of the same

kind. That is why the genotype-phenotype distinction is essential to any full definition and

explanation of a Lamarckian process. Those that think otherwise are challenged to provide a

definition of Lamarckism that uses the concept of inheritance in a sense that excludes

contagion.

The genes are not the only form of genotype, even in the biological sphere. Following many

other authors, we propose that genotypes (or replicators) exist at the social as well as the

biological level. Candidates include ideas, memes, habits and routines. The possibility of

Lamarckism at the socio-economic level hinges on the existence of two mechanisms: one that

encodes acquired phenotypic characteristics in the genotype and another that conveys the

acquired characteristics from social genotype to social genotype. We examine this possibility

below.

- 6 -

This paper is organised in four sections. The second section concerns biological evolution

and the theoretical reasons why any Lamarckian transmission must be limited (if indeed it

exists at all) in any biological system. Several of these theoretical reasons turn out to have a

broader applicability than to biology alone. The general significance of the genotype-

phenotype distinction is also established. We show this by a visit to an imaginary planet

where Lamarckian transmission does exist among its biological species. This sets the stage for

the discussion of socio-economic evolution in the third section. Some possible social

genotypes are considered, with a view to exploring possible distinctions between genotype

and phenotype at the social level. We then consider the possibility of, and limits to,

Lamarckian transmission in socio-economic evolution. The fourth section concludes the

essay.

2. Problems on Planet Lamarck

To explore further the theoretical limits to Lamarckism, we consider the viability of a

hypothetical inheritance system on Planet Lamarck. We shall later explain the significance of

this discussion for socio-economic evolution on Planet Earth. The first humans to explore

Planet Lamarck were a group of evolutionary economists and organisation theorists.7 They

observed the reproduction of several species, including a giraffe-like organism. They noted

that with each generation some characteristics would be exaggerated. With the giraffes, for

example, each generation would give rise to offspring with a longer neck, resulting in a

discernable increase in neck length through the giraffe lineage, from generation to generation.

One evolutionary economist quoted Joseph Schumpeter and argued that this evolutionary

process seemed to operate in the giraffe species as if „from within‟. But her colleagues

pointed out that the increase of neck length occurred in an environment where the giraffes

depended on sustenance and were reaching for the uppermost leaves in the trees, so the

impact of the environment should not be ignored. They persuaded each other that biological

evolution on this planet was in fact Lamarckian, and similar in its essentials to the

Lamarckian processes of socio-economic and cultural evolution that they had analysed in

human society on Earth. Accordingly they named the planet „Lamarck‟.

7 Our lawyers have urged us to state that any resemblance to any living evolutionary economist or organisation

theorist is purely coincidental.

- 7 -

Generation 1 Generation 2 Generation 3

Environment E1 E2 E3

Organism O1 O1' O2 O2' O3 O3'

Figure 1: A Process of Evolution (or Contagion?)

In their report on the biological evolution of organisms on this planet the social scientists

included the diagram in figure 1. Their account of the evolution of organisms on Planet

Lamarck noted the individual developmental process of each organism from Oi to Oi' in each

generation i, subject to environmental influences Ei. The offspring in the next generation

inherited and started with the acquired characteristics of Oi' in the form Oi+1 and these were in

turn developed and augmented into Oi+1'. This process was then repeated indefinitely. The

observers proposed that this was formally similar to their Lamarckian models of learning

processes in human organizations and cultures on Earth, where in each discrete stage,

knowledge Ki builds on and develops to Ki' while adapting and testing in environmental

conditions Ei, and this knowledge is accumulated and transmitted onto the next stage.

The social scientists added a caveat in their report that they did not have training in biology

and it would be necessary for a group of evolutionary biologists to explore Planet Lamarck in

order to confirm their observations and results. The evolutionary biologists on Earth were

very critical of the report. They asked: how could it be possible for a characteristic acquired in

the development of one organism to be passed on to the next generation? The next generation

is not a mere photocopy of its predecessor, so what mechanism could account for the

transmission of these characteristics from generation to generation?

In addition, the evolutionary biologists on Earth pointed out that the model portrayed in

figure 1 cannot distinguish between genuine inheritance and virus-like contagion. This

distinction again depends on attention to the mechanisms of inheritance that were treated

inadequately in the report by the social scientists.

The critics also objected that this picture of „Lamarckian‟ inheritance does not explain why

„acquired‟ improvements are favoured over acquired impairments or injuries. If an organism

becomes aged or infirm, or is injured or mutilated, then no reason is given why these

- 8 -

impairments are not immediately passed on to the offspring. Presumably, these characteristics

would also be apparent at their birth. So newborn giraffes would not only have necks as long

as their parents, but would also inherit any rheumatism, diminished virility and failing

eyesight. The newborn giraffe would have the same neck length as its mother, and it would

not fit into her womb. If valid, Lamarckism cannot entail that the next generation starts where

the previous generation ends.

In their application for a research grant to finance a second exploration of Planet Lamarck,

the evolutionary biologists hypothesised for these reasons that it was unlikely that all acquired

characters would be inherited. There must be some mechanism that prevented this.

They also set out a framework using the biological concepts of genotype and phenotype.

They noted that in particular the evolutionary economists had made little use of these key

concepts, but these were necessary to sustain an adequate account of Lamarckian evolution on

Planet Lamarck or elsewhere, if indeed it existed. They also noted that other theorists of social

or cultural evolution, including enthusiasts of „memes‟, had either failed to mention the

genotype-phenotype distinction in that context, or failed to reach a consensus in the

identification and consistent specification of the meme-genotype or the meme-phenotype

(Dawkins, 1976; Blackmore, 1999).8

Pouring more scepticism on the claims of the memeticists and other social scientists, the

evolutionary biologists cited a paper by David Hull (1982) that argues that memetic evolution

could not be Lamarckian, but must be Darwinian. For Hull (1982, p. 278) „social learning is

not an instance of the inheritance of acquired characters‟. For him, it is more like

epidemiological infection or contagion. He thus rejects the notion that Lamarckian

transmission is involved. For Hull (1982, p. 309), the inheritance of acquired ideas or memes

is not an instance of the inheritance of acquired characters, because ideas and memes are

analogous to genes, not characteristics. „In order for sociocultural evolution to be Lamarckian

8 Neither Lamarck nor Darwin used the terms genotype or phenotype, but some such conceptual separation was

implicit in their discourse. Writing in a biological context, they both assumed that information related to

characteristics was transmitted from generation to generation through such information-carrying entities as

seeds, sperm, ova, pollen, or stigma. When Darwin (1859, 1868) wrote of the „reproductive system‟ and of „germ

cells‟ he was referring to the processes of genotypic, not phenotypic, replication. Neither writer believed that

replication occurred by the direct copying of characteristics, so that their writing in the biological context

embodied an implicit distinction between genotype and phenotype.

- 9 -

in a metaphorical sense, conceptual genotypes must be distinguishable from conceptual

phenotypes and the two must be related in appropriate ways.‟ If we make this important

distinction, and if we choose to treat memes or ideas as genotypes, then the spreading of ideas

or memes is like the spreading of organisms, and does not necessarily involve the inheritance

of acquired characters.

Social learning and cultural transmission can be other than the contagious spread of ideas,

but only if meaningful social genotypes (or replicators) exist. We believe they do, as

explained below. A key point made by Hull is that the transmission of ideas or memes is the

spread of entities that are closer to genotypes not characteristics. If we deny this, or regard the

genotype-phenotype distinction as unwarranted, then we have no way of distinguishing

between acquired character inheritance and contagion.

In emphasising the importance of the genotype-phenotype distinction, the evolutionary

biologists expanded the crude evolutionary picture in figure 1, into the more sophisticated

presentation in figure 2. In both cases, the diagram omits a selection effect in order to focus

on Lamarckian inheritance. The presence of selection would add an effect at the population

level where entities with less fit phenotypes would exit the population and, as a consequence

of their exit, the distribution of genotypes would change (because their genotype might be

eliminated from the population). New entities would enter as a consequence of replication,

with the more fit phenotypes leaving more replicas than the less fit. This would further alter

the distribution of genotypes. In the presence of a fairly stable environmental factor and a

replication process that reliably transmitted genotypes between generations, the population

would slowly and systematically adapt to the environmental factor. The biologists knew all

this, but following good scientific practice, they wanted to examine the effect of Lamarckian

inheritance in isolation.

Generation 1 Generation 2 Generation 3

Environment E1 E2 E3

Phenotype P1 P2 P3

d1 l1 d2 l2 d3 l3

Genotype G1 G1' G2 G2' G3 G3'

Figure 2: An Expanded Evolutionary Schema with Lamarckian Inheritance

- 10 -

Figure 2 illustrates the supposed inheritance process on Planet Lamarck, with its reported

inheritance of acquired characters.9 The symbols Gi and, Gi′ refer to the genotypes of

organism i in a lineage of organisms. Pi refers to the developed phenotype. In order to keep

the presentation simple, we have not distinguished between stages of phenotypic development

in the figure. It must be emphasised that phenotypic development is an outcome of both the

genotype and the environment. In the construction of a more complete evolutionary theory,

developmental process must be brought fully into the picture.

The first organism has a genotype G1 that instructs its development d1 (denoted by a thicker

arrow) into phenotype P1. This phenotype reflects environmental conditions E1 as well as

genotypic characteristics G1. Somehow, through a process of Lamarckian inheritance l1

(denoted by a down-pointing arrow), some or all of the characteristics of this phenotype are

encoded in the same organism‟s genotype, so G1 transforms into G1'. Of course, such

Lamarckian inheritance l is minimal or absent at the biological level on Earth. But on Planet

Lamarck, we can consider the possibility that a fraction of information might be transferred

with an organism from phenotype to genotype. Some or all of this information might be

transferred in process l.

Next, through mating or whatever, replication occurs. So the information in genotype G1' is

passed on to the next generation in the form of genotype G2. The same process repeats in this

and subsequent generations. By the time we have reached the third generation, the genotypic

outcome G3' carries information gathered from its ancestral genotypes, including some

accumulated phenotypic information encoded in genotypes, reflecting previous environmental

conditions E1, E2 and E3.

The evolutionary biologists noted that without the conceptual distinction between genotype

and phenotype, the „phenotype‟ and the „genotype‟ rows in figure 2 would be conflated into a

single row, and the characteristically Lamarckian process l would disappear from the picture.

Accordingly, claims that (biological or socio-economic) evolution is Lamarckian, whether

valid or not, depend on a clear distinction between genotype and phenotype in order to be

adequately meaningful. Relatively few of the previous advocates of Lamarckian socio-

9 Similar multi-level transmission diagrams can be found in Boyd and Richerson (1985) and Durham (1991), and

are now in widespread use in the literature.

- 11 -

economic evolution have paid sufficient attention to this point. This is curious because the

very existence of Lamarckian transmission turns on the phenotype-genotype distinction

(Aunger, 2002; Hull, 1982, 1988, 2000).

Some Lamarckians have denied that developments such as the growth of a giraffe‟s neck

must result from genotypic instructions. It is here that the third and „volitional‟ interpretation

of Lamarckism is sometimes invoked. But even if the giraffe could use its own will power to

stretch its neck, we would then have to explain why the giraffe has a disposition to act

purposefully in this manner.10 The cause and evolutionary origin of this volitional propensity

would itself have to be explained. No answer to this question is available except for a

Darwinian one, in which such a wilful propensity somehow gives the giraffe a fitness

advantage and it is thus favoured by natural selection. But this argument also requires that the

volitional propensity is itself genetically encoded. Consequently, the propensity to stretch the

neck again derives from the biological genotype.

Lamarck ([1809], 1984, p. 113) himself argued that „a more frequent and continuous use of

any organ gradually strengthens, develops and enlarges that organ … while the permanent

disuse of an organ imperceptibly weakens and deteriorates it‟. This famous Lamarckian

principle of „use and disuse‟ does not help matters either. Again we must search for a causal

explanation, why the use of an organ leads to enlargement or strengthening, and disuse to

diminution. For these processes to occur in a systematic way, there must be a mechanism in

the body that reacts to use or disuse, and causes such strengthening or weakening. This

mechanism must be inherited, and thus must be an outcome of the genotype. Hence

Lamarck‟s principle of „use and disuse‟ cannot escape the requirement that the outcomes

derive from instructions in the genotype played out in specific environmental conditions.

In systems without Lamarckian inheritance, the environment impacts on a distribution of

genotypes solely through a process of selection on the expressed phenotypic properties (traits)

present in the whole population. Some phenotypes are less adapted than others to a given

environment. Given a systematic relationship between genotypes and phenotypes, the gene

10 Additionally, we would also have to explain the presence of the level of phenotypic plasticity defining the

giraffe‟s behavioural options (how much can it choose to stretch its neck?). In principle, nothing in the

Lamarckian argument precludes that the giraffe would be able to change the level of phenotypic plasticity

through volition. However, accepting this argument clearly exposes the untenable logic of Lamarckism.

- 12 -

pool can thus change from generation to generation as a result of selection (Price, 1995). On

Planet Lamarck, selection can occur, but by contrast it is no longer necessary to account for

the evolution of the genotype. Accordingly and historically, Lamarckians such as Herbert

Spencer gave relatively less emphasis to selection in their evolutionary theory.

In their grant application, the evolutionary biologists elaborated the following theoretical

argument. The inheritance of acquired characters has to be distinguished from something akin

to epidemiological infection or contagion, where one phenotype influences a second

phenotype without corresponding changes in the second genotype. A newborn giraffe‟s

propensity to grow a longish neck must be encoded in the genotype of the newborn giraffe.

Especially with Lamarckian transmission, this genotypic inheritance mechanism is necessary

to avoid newborn giraffes having necks as long as their parents. Instead these offspring inherit

a genotypic propensity to grow long necks. This genotypic propensity is passed from

generation to generation.

However, there is nothing specifically Lamarckian about the inheritance of a propensity to

grow a long neck. Earthly giraffes inherit such a propensity, without Lamarckian meddling

with their DNA. For Lamarckian inheritance to occur, the longer necks of the parents must

further enhance the propensity to grow a long neck that is encoded in their genotype, and this

enhanced genotypic propensity must then be passed on to the genotype of their offspring.

Hence Lamarckian replication must involve the following two essential stages:

(L1) although its genotype already contains instructions to develop a particular

characteristic, (such as a long neck) this realised phenotypic outcome somehow causes an

amplification of these genotypic instructions, to enhance this characteristic even further;

and

(L2) through reproduction, the instructions that favour this additional development are also

passed on to the next generation.

Compare these points with the definition of Lamarckism by Hull (2000, pp. 55-6), who wrote

inheritance is Lamarckian if the environment changes the phenotype of an organism in

such a way that this organism is better adapted to the environmental factor that produced

this change. This phenotypic change must then be transmitted somehow to the genetic

material so that it can be passed on to the offspring of the organism through reproduction.

These offspring then are born with this acquired characteristic more highly developed or

- 13 -

with a strong tendency to produce this characteristic more highly developed. Lamarckian

inheritance is the literal inheritance of acquired characteristics. The transmission must be

genetic, and the relevant effect must be phenotypic.

Hull‟s formulation is close to ours.11 A key element in both L2 and Hull‟s statement requires

further emphasis. The Lamarckian „inheritance of acquired characteristics‟ must mean more

than the mere inheritance of the capacity to grow a long neck. The genotypic instructions that

lead to the realisation of this outcome must somehow lead to the amplification of those

instructions. Accordingly, Lamarckian organisms must have genotypes that provide positive

feedback on the genotypic instructions that promote this growth.

Having clarified the meaning of Lamarckian inheritance, the evolutionary biologists

produced some reasons for being sceptical of its existence. The first argument identified some

dangers in excessive feedback from phenotype to genotype. Given that particular

environmental cues trigger only a subset of a large range of phenotypic possibilities, an

organism represents only one of the adaptive outcomes that are possible given its genotype.

The genotype carries the accumulated wisdom of past generations, in many environments. To

preserve this valuable heritage, this genotypic „baseline‟ must not adjust too rapidly in

response to current phenotypic outcomes.12 Accordingly, Lamarckian inheritance would

somehow have to preserve much of the genetic material that is not actually expressed in the

current phenotype. Hence Lamarckian inheritance cannot be so strong that it distorts or

overwhelms this legacy.

The problem of altering genetic instructions to correspond to phenotypic change gets even

worse if we examine the required mechanism of back-translation. Back-translation of an

11 Note, however, that by insisting that „Lamarckian transmission must be genetic‟ rather than more broadly

genotypic, Hull seems to immediately rule out the possibility that (literal?) Lamarckian transmission may exist

using social and other genotypes, which are not genes. Rather than excluding the possibility of Lamarckian

social evolution by an act of definition at the outset, we prefer to replace the restrictive term „genetic‟ by the

much broader terms „genotypic‟ or „genotype‟ and explore the Lamarckian possibility in that context, as we do

in the next section.

12 See Maynard Smith and Szathmáry (1999), Knudsen (2001, 2002a). DNA has a remarkable mechanism for

limiting mutations. Consisting of two strands, one strand can be checked against the other by an enzyme, and if

necessary repaired by other enzymes.

- 14 -

acquired trait requires an accurate identification and modification of the genes that correspond

to this and only this trait, such as a long neck. For example, error in back-translation of an

acquired long neck might instead promote a smaller neck or larger feet.

All this presumes that the environment acts like an expert computer software redesigner, as

if understanding the complex interconnections between each piece of coding and knowing

which instructions to preserve and which to modify. Additionally, it must be anticipated how

the environment interferes with genetic instructions as the new organism develops from

embryo to maturity. Such a degree of detailed, complicated and fortuitous reprogramming is

very unlikely to happen in the haphazard and undesigned turmoil of nature. Readily solving

the problem of back-translating phenotypic traits to genetic information requires a one-to-one

mapping of phenotype onto genotype. No such mapping is known in biology or elsewhere.13

Another problem concerns the very meaning of an acquired character. The evolutionary

biologists pointed out that logically there were two entirely different types of acquired

characters, depending on whether it (i) resulted, or (ii) did not result, from instructions in the

genotype. In the first case, there is nothing specifically Lamarckian about a character resulting

from its genotypic coding. As argued above, Lamarckism must refer to the additional process

by which the genotype somehow encodes the characteristic in its enhanced form, so that the

next generation does not have to start developing from the same starting point.

In the second case, the „acquired character‟ does not result from instructions in the

genotype. Neither can it result from (unintentional or intentional) behaviour in the organism,

because behavioural dispositions may themselves be genetically inherited. The remaining

possibility is an accidental impact of the environment. Unfortunately, however, in any

realistic case most accidental impacts result in injuries. Hence the most straightforward case

of an acquired characteristic is an injury. But for species to evolve, such injuries must be

limited. Hence the effects of both types of „acquired character‟ must be highly limited for

evolution and adaptation to be sustained through time. To provide a complete explanation, we

need to account for the existence of sufficiently tight limits that disallow inheritance of

useless and injurious characters. The only possible explanation for the evolution of the limits

that are required to filter the many useless and injurious characteristics is natural selection.

13 Recent work in biology implies that the mapping of phenotype onto genotype is many-to-one or even many-

to-many (Stadler et al., 2001).

- 15 -

Accordingly, Lamarckism depends on the Darwinian principle of selection in order to explain

why any disastrous propensity to inherit acquired impairments does not prevail. As Richard

Dawkins (1986, p. 300) argues, „the Lamarckian theory can explain adaptive improvement in

evolution only by, as it were, riding on the back of the Darwinian theory.‟ Lamarckism, if

valid in any particular domain, depends on Darwinian mechanisms of selection for

evolutionary guidance.

Lamarckian inheritance requires natural selection for guidance, hence it must not over-

reach the effects of natural selection. We know that natural selection works very slowly and

erratically, so that the generational effects of any Lamarckian inheritance, with its strong

injurious bias, must be small by comparison.

Consider if there were competition on Planet Lamarck between Lamarckian and non-

Lamarckian species. For Lamarckian inheritance to be prevalent it must bestow an advantage.

But much acquired character inheritance is disadvantageous or injurious. The most obvious

advantage for Lamarckian inheritance is that it may allow the enhancement of fortuitous

adaptations to a given environment. However, this advantage would be reduced in complex or

changing environments, because Lamarckism might lead to the species being too quickly

locked in to an inferior peak in the fitness surface. Accordingly there are good theoretical

reasons, as on Earth, why biological evolution is largely or entirely non-Lamarckian.

Hence Lamarckism faces severe theoretical limits that may be derived independently of the

empirical investigation into the inheritance process. The evolutionary biologists made all

these theoretical points in their research grant application, and then made the case that

empirical enquiry into the inheritance mechanisms on Planet Lamarck was essential to

ascertain whether such an unlikely outcome of Lamarckian inheritance did in fact exist. They

left for Planet Lamarck some time ago, and we have yet to receive their research report.

In the meantime, it was noticed by some social scientists that many of the key points in the

application for research funding for the journey to Planet Lamarck were quite general, and

applied to all cases of evolution with a population of replicating entities. Accordingly, the key

theoretical observations of the evolutionary biologists, as outlined in this section, applied also

to socio-economic or cultural evolution on Earth, which several other authors had claimed to

be essentially Lamarckian. It is the purpose of the next section to report the implications of

this wide-ranging observation.

- 16 -

3. Back to Socio-Economic Evolution on Planet Earth

One of the messages of the fable in the previous section is that the distinction between

genotype and phenotype is vital to understand the mechanisms of an evolutionary process

involving inheritance and replication. Any specification of a fully-fledged evolutionary

process, involving a population of developing and replicating entities, must clearly identify

the relevant genotypes and phenotypes. This is quite a general point, and applies to socio-

economic and cultural evolution in human society, as well as to biological systems. In

particular, any claim that socio-economic or cultural evolution is Lamarckian depends on the

genotype-phenotype distinction for its explication.14

Remarkably, however, despite frequent claims of Lamarckism in these literatures, works in

evolutionary economics, organisation theory, evolutionary anthropology and memetics

seldom make use of the genotype-phenotype distinction, or explore the nature of genotypes or

phenotypes in their domain.15 There is much loose discussion of the transmission of ideas,

beliefs, knowledge or memes, but relatively little dissection of the precise processes involved.

We now apply the key insights of the preceding section to evolutionary processes in

economies and societies. First we ask by what criteria a (biological or social) entity qualifies

as a genotype. There is relatively little discussion of this question in the literature. Dawkins

(1976) argued that a replicable genotype must have the characteristics of longevity, fecundity

and replicative fidelity. Our DNA replicates with a high degree of precision and with a low

probability of mutation. By contrast, in the social domain, no candidate social genotype gets

close to DNA by Dawkins‟s criteria. We have to search for other distinguishing criteria.

These criteria must be able to identify social genotypes, which by definition are not

inherited biologically, which act as stores of social dispositions, rules and knowledge, and

which can guide the development of human patterns of behaviour and social structures,

depending on the overall context. Social genotypes are neither genes nor DNA; they are

14 However, Price‟s (1995) general formulation of the concept of selection does not make use of the genotype-

phenotype distinction (Knudsen, 2004). The distinction becomes important when the mechanisms of replication

and transmission are considered. The significance of the genotype-phenotype distinction was established in

Darwinian biology in the twentieth century.

15 Winter (1971), McKelvey (1982), Hull (2000) and Aunger (2002) are among the exceptions.

- 17 -

replicated by other means. A social genotype is a complex of dispositions to behave in

particular ways in particular situations. These behaviours may include thoughts as well as

actions.

Additional criteria become available when we consider the role of a genotype as a

replicator. The division between replicator and interactor closely parallels the distinction

between genotype and phenotype. As Robert Brandon (1996, p. 125) put it, the distinction

between replicators and interactors „is best seen as a generalization of the traditional

genotype-phenotype distinction‟. Accordingly, using the two distinctions in conjunction,

further highly general criteria become available. In his application of Darwinian principles to

the evolution of scientific knowledge, Hull (1988, p. 408) usefully defines a replicator as „an

entity that passes on its structure largely intact in successive replications‟ and an interactor as

„an entity that directly interacts as a cohesive whole with its environment in such a way that

this interaction causes replication to be differential.‟

Furthermore, detailed and general definitions of replication have been developed by Peter

Godfrey-Smith (2000), Dan Sperber (2000), and Robert Aunger (2002). There are slight

differences between their formulations, but all agree on the centrality of the following

necessary criteria for replication. (1) Causation: the source must be causally involved in the

production of the copy. (2) Similarity: the copy must be like its source in relevant respects.

And (3) Information transfer: the process that generates the copy must obtain the information

that makes the copy similar to its source from that same source.

Our next task is to locate some possible social genotypes. A number of candidates can be

put forward, including (a) ideas, (b) memes, (c) habits of thought or behaviour, or (d)

organizational routines. These are not necessarily mutually exclusive, and other possible

social genotypes may exist. Several memeticists have proposed that the „meme‟ is a general

concept that covers all other possibilities. More detailed analyses of the characteristics and

mechanisms regarding (c) and (d) are presented elsewhere, in which organizational routines

depend on individual habits as a substrate.16

16 See Hodgson (2003), Hodgson and Knudsen (2004), Knudsen (2002a, 2002b, 2004).

- 18 -

Ideas or Memes as Social Genotypes?

First we consider ideas and memes as possible genotypes. We noted above that there has been

some apparent difficulty in enforcing a distinction between genotype and phenotype in the

case of ideas and memes. With regard to ideas, can these be (genotypic) instructions that drive

behaviour, or (phenotypic) rationalisations of preceding actions or attitudes, or both?

Similarly, are ideas or memes replicators or interactors, or both? Without answers to these

questions, there is no possibility of adjudicating on the question of whether memetic evolution

is Lamarckian or not. Describing such evolution as Lamarckian would at best be ungrounded

and premature.

Some meme enthusiasts wish to retain a broad definition of the meme, and end up treating

the meme as a genotype in one context, and a phenotype in another. This creates havoc with

discussions of whether or not memetic transmission is Lamarckian. Susan Blackmore (1999,

pp. 61-2) proposes that whether memetic evolution is Lamarckian or not depends on whether

it is meme-as-behaviour or meme-as-instructions that is being copied. She argues that

copying-the-product brings the possibility of Lamarckian inheritance of acquired

modifications to the outcome, whereas copying-the-instructions does not; any alterations in

behaviour or outcome will not be passed on, because it is the instructions, not the outcomes,

that are being replicated.

However, even if we regard the meme-as-behaviour, and consider the copying of outcomes,

it is still misguided to jump to the conclusion that a Lamarckian possibility exists. In

particular, the identity of the genotype behind the (phenotypic) meme-as-behaviour is unclear.

If this genotype is literally the genes, then the analysis switches back to biological mode,

where we know that Lamarckian transmission is largely or entirely ruled out by the

Weismann barrier. Alternatively, for those that treat the meme as behaviour, there must be

some social genotype that corresponds to that behavioural phenotype. But this remains

unidentified. Consequently, within the meme-as-behaviour version of memetics, no case for

regarding memetic transmission as Lamarckian has been adequately established.

Some prominent proponents of the meme treat it as a genotype and a replicator, rather than

a phenotypic phenomenon, such as behaviour (Aunger, 2002; Hull, 1982, 2000). Hull

convincingly argues that memetic transmission, where memes are treated as ideas and

genotypes, cannot legitimately be described (literally or metaphorically) as Lamarckian. Hull

(1982, p. 311; 2000, p. 87) starts from the view that „memes are analogous to genes, not

- 19 -

characteristics‟. Consequently, memes may be modified or acquired, but this is neither the

modification nor the acquisition of a characteristic. Memetic transmission is the inheritance of

acquired memes, and memes are genotypes not phenotypic characteristics. Accordingly,

social learning and other forms of memetic transmission cannot be instances of the inheritance

of acquired characters. For Hull, the replication and spread of memes is more like

epidemiological infection or contagion. In conclusion, if we treat the meme as a genotype and

a replicator, then any description of memetic transmission as Lamarckian is mistaken.

Habits or Routines as Social Genotypes?

Let us now explore the possibility of habits and routines as social genotypes. In a series of

essays (cited in the preceding footnote) we have explored the possibility of regarding habits

and routines as replicators (or genotypes). This strategy avoids some of the vagueness and

difficulties associated with the „meme‟, but creates further problems for the use of the

„Lamarckian‟ label in the socio-economic context, as we shall explain below. We treat habits

and routines as dispositions, rather than expressed behaviour as such. If we acquire a habit we

do not necessarily use it all the time. It is a propensity to behave in a particular way in a

particular class of situations.

Similarly, we have argued that it is also preferable to treat routines as propensities

(Hodgson, 2003; Knudsen, 2002a). As Barbara Levitt and James March (1988, p. 320) put it:

„The generic term “routines” includes the forms, rules, procedures, conventions, strategies,

and technologies around which organizations are constructed and through which they

operate.‟ Michael Cohen et al. (1996, p. 683) concur in treating a routine as a disposition: „A

routine is an executable capability for repeated performance in some context that [has] been

learned by an organization in response to selective pressures.‟

In evolutionary, developmental and functional terms, instinct comes before habit, and habit

comes before belief and reason. Instincts provide inherited behavioural cues that guide us

initially in our newborn state. Then our actions, resulting from instinct or cultural interaction,

lead to the formation of habitual dispositions. In turn, these habits or dispositions form the

basis of our conceptualisations and beliefs. Thus, habits are the basis of both reflective and

nonreflective behaviour (Kilpinen, 1999, 2000; Hodgson, 2004).

The evolution of instincts, which by definition are inherited genetically, is a matter for

biology. We know that Lamarckian transmission is essentially ruled out in biology, and

- 20 -

human instincts evolve very slowly. By contrast, habits are largely a cultural phenomenon and

can evolve much more rapidly. They are dispositions that are acquired through repeated

mental or physical behaviours, in specific social contexts. Habits can sometimes spread

through a culture, resulting in a remarkable uniformity of dispositions and eventually beliefs,

even in a large society.

Can Habit Replication be Described as Lamarckian?

Our theoretical strategy is to treat habits as genotypes, in the sense that they preserve and

transmit social roles, attitudes, knowledge and skills, and act as the relatively durable

substrate of all beliefs and deliberative reason. To establish this point fully would be to divert

us from the main theme of this essay, but our reasons are laid out more extensively elsewhere,

as well as in the pragmatist literature as a whole.17

An important corollary is worth raising at this stage. From the biological point of view,

habits are part of the biological phenotype; they are expressions of genetic instructions in

interaction with the environment. However, from the social viewpoint, habits become

genotypes. This is partly because they satisfy the conditions of longevity, fidelity and

fecundity, relative to the shorter time-scales and wider margins at the social level. An aspect

of the phenotype at one level becomes a potential genotype at a higher level, but under

different standards of longevity and fidelity. Although the relevant criteria to establish such

classifications are not yet sufficiently refined, the general idea of phenotypic features acting

as genotypes at a higher level is necessary for a multiple level selection theory. Accordingly,

habits are both phenotypes (with regard to genetic replicators) and genotypes (with regard to

socio-economic evolution).

Addressing the replication of habits, first consider habits relating to observable behaviour.

Unlike the replication of DNA or computer viruses, habits of behaviour do not directly make

copies of themselves. Instead they replicate indirectly, by means of their behavioural

expressions. They can impel behaviour that is consciously or unconsciously followed by

others, as a result of incentive or imitation. It is possible, but not always necessary, that

codifiable rules or instructions are also involved. Eventually, the copied behaviour becomes

17 See Hodgson (2004), Joas (1996) and Kilpinen (2002), as well as classics such as Dewey (1922) and Veblen

(1914), and modern psychological approaches such as Oullette and Wood (1998) and Wood et al. (2002).

- 21 -

rooted in the habits of the follower, thus transmitting from individual to individual an

imperfect copy of each habit by an indirect route. Referring to the definition of replication by

Godfrey-Smith, Sperber and Robert Aunger noted above, in what sense is a replicated habit of

behaviour similar to the one from which it derived? The similarity results from the shared

propensity to perform similar behaviours under similar conditions.

Often we repeatedly follow the behaviour of others, because of an instinctive or acquired

disposition to conform. Behaviour can also be moulded by incentives or constraints, giving

reasons to acquire specific customs, follow particular traffic conventions and use specific

linguistic terms. In following others, we too build up habits associated with these behaviours.

The behaviours are imitated and also the habits giving rise to them are replicated.

Habit replication often relies on imitation. Imitation need not be fully conscious, and it will

also involve some „tacit learning‟ (Polanyi, 1967; Reber, 1993; Knudsen, 2002a). Perhaps

imitation can occur even without strong incentives, on the grounds that the propensity to

imitate is instinctive, and this instinct has itself evolved for efficacious reasons among social

creatures (Campbell, 1975; Boyd and Richerson, 1985; Simon, 1990; Tomasello, 2000).

However, an imitation instinct might require an existing set of common behaviours in the

group, otherwise an emerging propensity to imitate might not have a selection advantage.

Furthermore, if imitation is more than mimicry, then the rules and understandings associated

with it also have to be transmitted.

Because habits of thought are unobservable, they cannot replicate straightforwardly via

imitation. To show how habits of thought are replicated, we need to consider how thoughts

are limited by constraints that somehow bear upon mental activity. The replication of habits of

thought is guided by similar external constraints and similar inborn constraints of brain

design. However, such replication also requires the communication of similar mental models,

so that two or more individuals may interpret given sense data in a similar way. Hence this

replication process cannot be completed without the existence of a common language. The

ability to refer to common mental models through language enables communication and

thereby replication of mental models.

Of course, we must have some other prior instincts or habits that enable us to recognise

incentives or imitate others. Habit replication thus itself depends on prior instincts or habits.

Some kind of (social or biological) genotype has to exist for the behaviour to be imitated. It

may operate via the copying of (phenotypic) behaviours, but always relies on prior genotypic

- 22 -

instructions of some kind. Habit replication via a phenotype implies neither the absence nor

unimportance of a genotype, at any stage of the process.

Can habit replication be Lamarckian? Hodgson (2001, 2003) suggested this elsewhere, even

if this possibility is highly limited. Because the replication of some habits works through the

phenotypic and behavioural level, any additional behavioural characteristic that did not relate

to the original habit might also be transmitted to the receiver. With some habits, acquired

characters can be inherited because their replication works through characteristics, not

through the direct replication of the generative structures.

We now wish to refine this assessment of habit replication as potentially Lamarckian. We

show below that it all depends on how we translate the Lamarckian steps L1 and L2 above

from biological to social terms. The transmission of habitual genotypes is always indirect.

Lamarckism would have to be rendered consistent with indirect transmission, yet the work of

Lamarck himself, and much of relevant subsequent literature, relates to the biological world,

where genotypic transmission is always direct rather than indirect. It is an important question

(revisited below) whether indirect transmission should in principle be admitted within the

Lamarckian schema.

Serious problems also arise with step L2. A Lamarckian process that is defined in these

terms would require that the relevant aspect of the phenotype (an acquired thought or

behaviour) of the first person was also back-translated into its genotype (habit). This may

occur when repeated (phenotypic) thoughts or behaviours give rise to new or amended

(genotypic) habits. But the phenotypic behaviour could be occasional or accidental and not

encoded in a habit, yet still be imitated by the second person. Strictly, according to the

formulation in L2, the first case would be Lamarckian but the second would not.

Environment E1 E2 E3

Behaviour (phenotype) B1 B1' B2 B2' B3 B3'

l1 l2 l3

Habit (genotype) H1 H1' H2 H2' H3 H3'

Figure 3: The Replication of Habits of Behaviour

- 23 -

At first sight, this may seem to resolve the issue: a Lamarckian possibility exists with

regard to the replication of habits, as long as the acquired behaviour gives rise to an enhanced

habit in the first person before the behaviour is imitated by a second person. In figure 3,

behaviour B1 becomes ingrained in habit H1. Practice and interactions with the environment

lead to modified behaviour B1' and modified habits H1'. The second agent copies this modified

behaviour, and the process is repeated. A „Lamarckian‟ process of habit transmission is shown

in figure 3 by the downwards-pointing arrows l indicating the affect of the behavioural

phenotype on the habitual genotype.

Comparing figures 2 and 3, there is an important difference in the role played by the

„Lamarckian‟ step l. As shown in figure 3, this step affects neither the genotype nor the

phenotype of the person that imitates the behaviour. The modification of habit to Hi', which

occurs just prior to the imitation, plays no role in the replication process or its result. The

significance of this observation is that the definitionally essential Lamarckian step l plays no

causal role in the replication process. The step leads into a causal cul-de-sac. What is crucial

to the description of Lamarckian replication plays no vital role in the „Lamarckian‟ replication

of habits in the social domain! This outcome results from the fact that with the replication of

habits, genotype-to-genotype transmission is indirect.

Can Routine Replication be Described as Lamarckian?

Let us now consider the possibility of treating routines as organizational level genotypes.

When routines are copied from organisation to organisation then this in part may involve the

adoption by the imitator of similar and explicit rules and procedures. Even if routine

replication consisted entirely of the „blueprint‟ copying of codifiable procedures, then a

Lamarckian description would still be inappropriate, for reasons similar to Hull‟s objection to

the idea that meme replication is Lamarckian, as discussed above. Blueprint transmission of

routines is the inheritance of genotypes. There is no inheritance of the additionally acquired

phenotypic characteristics of the performed routines.

On the basis of extensive research on the nature of routines we know, however, that much

of the know-how inherent in routines is tacit and often uncodifiable (Polanyi, 1967; Nelson,

and Winter, 1982; Hannan and Freeman, 1989; Cohen and Bacdayan, 1994; Cohen et. al.,

1996). The formation of routines operates via the creation of interlocked habits of individuals

in a team. Hence habit replication is a part of routine replication. Often the routine must be

- 24 -

observed and practised, because the transfer of blueprint information is not enough to

consolidate the routine. As a result, the same problems that emerged with a „Lamarckian‟

description of habit replication occur with the replication of routines.

There is an extensive debate within organisation science as to how changes in routines in

firms occur, and whether the evolution of organisations and routines is a „Lamarckian‟

process. Near one extreme is the view that rules and routines are difficult to alter within any

specific organisation, and that changes occur principally through the selection and elimination

of some organisations, rather than adaptations of routines within the organisations themselves

(Hannan and Freeman, 1989). A huge case study literature, too massive to begin citing here,

testifies to the conservative nature of organisations, and the durability of their habits and

routines, even when more productive or efficient alternatives exist. However, our purpose

here is not to adjudicate over the empirical claims but to consider the applicability of the

„Lamarckian‟ label to significant adaptation and change in the routines in any given

organisation. The unfortunate fact is that the dispute within organisation science over the

extent to which routines can adapt is typically described as a contest between „Darwinian‟ and

„Lamarckian‟ conceptions of organisational change (Usher and Evans, 1996).

In part, what is problematic is the broad use of the „Lamarckian‟ label simply to describe

the adaptation of routines within any given organisation. Such descriptions have nothing to do

with the inheritance of characteristics (acquired or otherwise) from one organisation to

another. The use of the Lamarckian label to categorise a process that does not explicitly

involve inheritance is in defiance of most historical uses of the term.

In sum, we have explored several options for possible replicators in socio-economic

evolution. Every one carries problems for the application of the „Lamarckian‟ label in this

domain. If the genotype-phenotype distinction cannot be applied, then the Lamarckian

description is not meaningful. If it can be applied, then further problems arise. In the case of

memes-as-genotypes the further problem is that memes may be modified or acquired, but this

is neither the modification nor the acquisition of a characteristic. In the case of habits and

routines, a crucial and defining Lamarckian step plays no causal role in the replication

process.

The conclusion we draw is that Lamarckian concepts do not readily and meaningfully

transfer from biological to social evolution. But, by contrast, we have found no similar barrier

to the application of generalized Darwinian principles to the social domain. Darwinian

- 25 -

concepts can be generalised more readily, to cover all evolving systems with replicating

populations of some kind.

4. Concluding Remarks

The Lamarckian possibility has a curious attraction in the social sciences. Despite there being

scant evidence and inadequate conceptualisation of the actual mechanisms of replication in

the social world, many argue that socio-economic processes are Lamarckian. But why would

anyone favour the Lamarckian explanation over the Darwinian, given the difficulties and

incompleteness of the Lamarckian explanation?

We have noted that enthusiasts of the Lamarckian label often emphasise the substantial

transmission of knowledge and skill from one generation to another in socio-economic

evolution, and that nothing like this occurs among other species. Our response is that this

process does not involve the inheritance of acquired characters, in an adequate sense that

excludes virus-like contagion.

Consider the difference between the spread of fashions in clothing, and the contagious

spread of laughter around a room. In the latter case, as Dan Sperber (2000) argues, there is

copying of laughing behaviour but without the copying of the capacity to laugh. There is no

inheritance of this capacity: the ability to laugh already exists and it is triggered by the

stimulus. Without such inheritance this process cannot be Lamarckian.

By contrast, the spread of a fashion can often lead to a change of individual preferences,

habits or dispositions. If so, then the disposition to wear or admire particular styles of clothing

is inherited. The Lamarckian description passes the inheritance test, but it is only possible by

making a distinction between behaviour and dispositions, close to the distinction between

phenotype and genotype. Furthermore, as shown above, Lamarckian descriptions are

unwarranted unless the acquired behaviour gives rise to an enhanced disposition in the first

person before the behaviour is imitated by a second person. However, this enhanced

disposition plays no role in the replication process or its result! The Lamarckian description

may be strictly valid in this limited case only, but it is misleading because it illuminates no

vital step in the inheritance process.

Explaining social evolution requires a valid inheritance model, one that identifies the

underlying cause of the material that is transmitted among agents. Otherwise, there is no way

- 26 -

of knowing whether observed changes are outcomes of selection processes, drift, or

something else. Once we look at these processes, we can see that the inheritance of acquired

characters is at most highly limited even in the socio-economic sphere.

Of course, the mechanisms of evolution in the social and the biological spheres are very

different. But the irony, as we have shown here, is that the essential Lamarckian principles are

more closely and exclusively related to the biological sphere than they are to the social,

notwithstanding the fact that Lamarckian replication is rare or nonexistent with biological

species on Planet Earth. The very concept of Lamarckism depends on relatively direct

genotype-to-genotype replication, and a clear genotype-phenotype distinction. Prominent

accounts of the replicators in socio-economic evolution lack one or both of these elements.

Consequently, any use of the Lamarckian label to differentiate social from biological

evolution would be misplaced and misconceived.

By contrast, the core Darwinian principles of variation, replication and selection seem to

have a more general applicability, and evidently would apply to social as well as biological

evolution. Those that insist that social evolution is Lamarckian often ignore the more

accommodating nature of general Darwinian principles.

Against this, some organisation theorists propose that volition, deliberation, purpose,

planning and learning cannot be explained by the simple primitives of the general Darwinian

principles. For example, Joel Baum and Paul Ingram (1998) argue that organizations capable

of adapting during their life-time are more Lamarckian than Darwinian. Similarly, David

Rigby and Jurgen Essletzbichler (1997) and others, view change in firms resulting from

profit-induced search, learning and imitation as Lamarckian. In particular, it has been argued

that Lamarckian evolution promotes quick learning (Bruderer and Singh, 1996; Nelson and

Winter, 1982).

These arguments are representative of a common and persistent usage of the „Lamarckian‟

label in the literature on economic and organizational evolution. According to van de Ven and

Poole (1995), those adopting Darwinian evolution of social organizations argue that traits are

inherited through intergenerational processes whereas those following Lamarck argue that

traits are acquired within a generation through learning and imitation. It is surprising to see

that many works actually agree with this distinction and promote Lamarckian evolution over

Darwinian (Hedlund, 1994; Helfat, 1995; Metcalfe, 1994; Rosenberg, 1992) even without

much consideration of the nature of either form of evolution.

- 27 -

These arguments are problematic for a number of reasons. First, as shown by our

definitions above, Lamarckism and Darwinism are not in principle mutually exclusive.

Second, in order to speak of Lamarckian evolution, traits that are acquired within a generation

must also be inherited through intergenerational processes. The fast and loose

conceptualisation of Darwinism and Lamarckism that is widespread in much of the literature

on organisations is unfortunate because it uses a simplistic and insufficient distinction.

Acquiring traits through learning and adaptation is a necessary condition of Lamarckism, but

it is not sufficient. In order to qualify as Lamarckism, the acquired traits must also be encoded

in a genotype that is passed on to the next generation. Third, in a Darwinian explanation by

natural selection, the traits of an entity develop according to the instructions in its genotype

and the influence of environmental conditions. The process of development opens the

possibility that the individual entity adapts to the environmental conditions within the

possibilities given by its instruction set. It is therefore important to understand how the

genotype maps onto the phenotype. Moreover, the instructions for a character may be quite

open-ended, allowing multiple conditional responses or a gradual fixation through learning.

Rather than merely distinguishing between entities on the basis of the range of their

behavioural options, we must carefully examine the underlying mechanisms of the

transmission of such options.

Fourth, as regards the issue of Lamarckism promoting quick learning, it is unclear what is

meant. Lamarckism surely promotes a quicker encoding of the properties of the environment,

at least if we assume that acquired traits are somehow correctly back-translated into the

genotype. However, this quicker encoding may have little effect on the pace of learning

relative to the rate of replication. For example, viruses and bacteria reproduce themselves

much more rapidly than most, if not all, ideas in the social realm (Hull, 2000). In order to

understand quick learning at the population level, we need to take a closer look also on

replication strategies. And in order to understand individual level learning, we must examine

how the genotype of an entity maps onto its phenotype.

In order to demonstrate Lamarckian inheritance, a phenotype-genotype distinction must

first be established. Adaptations acquired at the phenotypic level must be encoded at the

genotypic level and then passed on to new generations. Unfortunately, in the literature on

economic and organizational evolution there is no clear conceptualisation of the phenotype-

- 28 -

genotype distinction. Without this conceptualisation there is no basis on which to adjudicate

whether economic and organizational evolution is Lamarckian or Darwinian.

More generally, we believe that it is necessary and useful to make a distinction between

social (i.e. non-genetic) genotypes and social phenotypes. Some primitive evolving systems

do not have anything corresponding to a genotype, which carries generative information

through time and somehow passes it from structure to structure. Nevertheless, more complex

evolving systems are likely to develop genotypes and mechanisms of genotypic transmission.

Otherwise, the phenotypes could soak up and transmit any environmental disturbance and

cumulate irrelevant and erroneous traits. Genotypes provide a degree of baseline stability, so

that vaguely efficacious selection of phenotypes can occur. Such stability, exhibited in

substantial information storage and a high degree of cultural and institutional conservatism, is

also characteristic of socio-economic evolution. Without social genotypes, the cumulation of

every panic, craze, fad and fashion would completely disrupt the transmission and selection of

tried and tested social knowledge.

We have shown that with the specification of plausible social replicators or genotypes, the

Lamarckian description carries severe problems in the social domain. Whether we regard

ideas, memes, habits or routines as genotype and replicators, then the replication of each of

these cannot properly and usefully be described as Lamarckian. No such a priori problems of

transferability apply to the core Darwinian principles.

- 29 -

References

Aunger, Robert (2002) The Electric Meme: A New Theory of How We Think (New York: Free

Press).

Baum, Joel A. C. and Ingram, Paul (1998), „Survival Enhancing Learning in the Manhattan

Hotel Industry, 1898-1980‟, Management Science, 44(7), pp. 996-1016.

Boesiger, Ernest (1974) „Evolutionary Theories after Lamarck and Darwin‟ in Ayala,

Francisco J. and Dobzhansky, Theodosius (eds) (1974) Studies in the Philosophy of

Biology (London, Berkeley and Los Angeles: Macmillan and University of California

Press), pp. 21-44.

Blackmore, Susan (1999) The Meme Machine (Oxford: Oxford University Press).

Boyd, Robert and Richerson, Peter J. (1985) Culture and the Evolutionary Process (Chicago:

University of Chicago Press).

Brandon, Robert N. (1996) Concepts and Methods in Evolutionary Biology (Cambridge and

New York: Cambridge University Press).

Brandon, Robert N. (1999) „The Units of Selection Revisited: The Modules of Selection‟,

Biology and Philosophy, 14, pp. 167-180.

Bruderer, Erhard and Singh, Jitendra V. (1996) „Organizational Evolution, Learning, and

Selection: A Genetic-Algorithm-Based Model (in Special Research Forum: Organizational

Ecology)‟, Academy of Management Journal, 39(5), pp. 1322-1349.

Burkhardt, Richard W., Jr (1977) The Spirit of System: Lamarck and Evolutionary Biology

(Cambridge, MA: Harvard University Press).

Butler, Samuel (1878) Life and Habit (London: Trübner).

Campbell, Donald T. (1965) „Variation, Selection and Retention in Sociocultural Evolution‟,

in Barringer, H. R., Blanksten, G. I. and Mack, R. W. (eds) (1965) Social Change in

Developing Areas: A Reinterpretation of Evolutionary Theory (Cambridge, MA:

Schenkman), pp. 19-49. Reprinted in General Systems, 14, 1969, pp. 69-85.

- 30 -

Campbell, Donald T. (1975) „On the Conflicts Between Biological and Social Evolution and

Between Psychology and Moral Tradition‟, American Psychologist, 30(12), December, pp.

1103-26.

Cohen, Michael D. and Bacdayan, Paul (1994) „Organizational Routines are Stored as

Procedural Memory – Evidence from a Laboratory Study‟, Organization Science, 5(4),

November, pp. 554-68.

Cohen, Michael D., Burkhart, Roger, Dosi, Giovanni, Egidi, Massimo, Marengo, Luigi,

Warglien, Massimo, and Winter, Sidney (1996) „Routines and Other Recurring Action

Patterns of Organizations: Contemporary Research Issues‟, Industrial and Corporate

Change, 5(3), pp. 653-98.

Corning, Peter A. (1983) The Synergism Hypothesis: A Theory of Progressive Evolution (New

York: McGraw-Hill).

Darwin, Charles R. (1859) On the Origin of Species by Means of Natural Selection, or the

Preservation of Favoured Races in the Struggle for Life, 1st edn. (London: Murray).

Darwin, Charles R. (1868) The Variation of Animals and Plants Under Domestication, 2

vols., 1st edn. (London and New York: Murray and Orange Judd).

Darwin, Charles R. (1871) The Descent of Man, and Selection in Relation to Sex, 1st edn., 2

vols (London: Murray and New York: Hill).

Dawkins, Richard (1976) The Selfish Gene, 1st edn. (Oxford: Oxford University Press).

Dawkins, Richard (1983) „Universal Darwinism‟, in D. S. Bendall (ed.) (1983) Evolution

from Molecules to Man (Cambridge: Cambridge University Press), pp. 403-25.

Dawkins, Richard (1986) The Blind Watchmaker (Harlow: Longman).

Dewey, John (1922) Human Nature and Conduct: An Introduction to Social Psychology, 1st

edn. (New York: Holt).

Durham, William H. (1991) Coevolution: Genes, Culture, and Human Diversity (Stanford:

Stanford University Press).

Godfrey-Smith, Peter (2000) „The Replicator in Retrospect‟, Biology and Philosophy, 15, pp.

403-23.

- 31 -

Hannan, Michael T. and Freeman, John (1989) Organizational Ecology (Cambridge, MA:

Harvard University Press).

Hayek, Friedrich A. (1988) The Fatal Conceit: The Errors of Socialism. The Collected Works

of Friedrich August Hayek, Vol. I, ed. William W. Bartley III (London: Routledge).

Hedlund, Gunnar (1994) „A Model of Knowledge Management and the N-Form Corporation‟,

Strategic Management Journal, 15, pp. 73-90.

Helfat, Constance E. (1994) „Evolutionary Trajectories in Petroleum Firm R&D‟,

Management Science, 40(12), December, pp. 1720-1747.

Hirshleifer, Jack (1982) „Evolutionary Models in Economics and Law: Cooperation versus

Conflict Strategies‟, in R. O. Zerbe Jr and P. H. Rubin (eds), Research in Law and

Economics, 4, pp. 1-60.

Ho, Mae-Wan and Saunders, Peter T. (eds) (1984) Beyond Neo-Darwinism: An Introduction

to the New Evolutionary Paradigm (London: Academic Press).

Hodgson, Geoffrey M. (2001) „Is Social Evolution Lamarckian or Darwinian?‟ in Laurent,

John and Nightingale, John (eds) (2001) Darwinism and Evolutionary Economics

(Cheltenham: Edward Elgar), pp. 87-118.

Hodgson, Geoffrey M. (2002) „Darwinism in Economics: From Analogy to Ontology‟,

Journal of Evolutionary Economics, 12(2), June, pp. 259-81.

Hodgson, Geoffrey M. (2003) „The Mystery of the Routine: The Darwinian Destiny of An

Evolutionary Theory of Economic Change‟, Revue Économique, 54(2), Mars, pp. 355-84.

Hodgson, Geoffrey M. (2004) The Evolution of Institutional Economics: Agency, Structure

and Darwinism in American Institutionalism (London and New York: Routledge).

Hodgson, Geoffrey M. and Knudsen, Thorbjørn (2004) „The Firm as an Interactor: Firms as

Vehicles for Habits and Routines‟, Journal of Evolutionary Economics, 14(3), July, pp.

281-307..

Hodgson, Geoffrey M. and Knudsen, Thorbjørn (2006) „Why We Need a Generalized

Darwinism: And Why Generalized Darwinism is not Enough‟, Journal of Economic

Behavior and Organization (forthcoming).

- 32 -

Hull, David L. (1982) „The Naked Meme‟, in Henry C. Plotkin (ed.) (1982) Learning,

Development and Culture: Essays in Evolutionary Epistemology (New York: Wiley), pp.

273-327.

Hull, David L. (1988) Science as a Process: An Evolutionary Account of the Social and

Conceptual Development of Science (Chicago: University of Chicago Press).

Hull, David L. (2000) „Taking Memetics Seriously: Memetics Will Be What We Make It‟, in

Aunger, Robert (ed) (2000) Darwinizing Culture: The Status of Memetics as a Science

(Oxford and New York: Oxford University Press), pp. 43-68.

Jablonka, Eva, Lachmann, M. and Lamb, M. J. (1992) „Evidence, Mechanisms and Models

for the Inheritance of Acquired Characters‟, Journal of Theoretical Biology, 158, pp. 245-

68.

Joas, Hans (1996) The Creativity of Action (Chicago: University of Chicago Press).

Kilpinen, Erkki (1999) „What is Rationality? A New Reading of Veblen‟s Critique of

Utilitarian Hedonism‟, International Journal of Politics, Culture and Society, 13(2), pp.

187-206.

Kilpinen, Erkki (2000) The Enormous Fly-Wheel of Society: Pragmatism’s Habitual

Conception of Action and Social Theory (Helsinki: University of Helsinki).

Knudsen, Thorbjørn (2001) „Nesting Lamarckism within Darwinian Explanations: Necessity

in Economics and Possibility in Biology?‟, in Laurent, John and Nightingale, John (eds)

(2001) Darwinism and Evolutionary Economics (Cheltenham: Edward Elgar), pp. 121-59.

Knudsen, Thorbjørn (2002a) „The Significance of Tacit Knowledge in the Evolution of

Human Language‟, Selection, 3(1), pp. 93-112.

Knudsen, Thorbjørn (2002b) „Economic Selection Theory‟, Journal of Evolutionary

Economics, 12(3), September, pp. 443-70.

Knudsen, Thorbjørn (2004) „General Selection Theory and Economic Evolution: The Price

Equation and the Replicator/Interactor Distinction‟, Journal of Economic Methodology,

11(2), June, pp. 147-73.

Lamarck, Jean Baptiste de (1984) Zoological Philosophy: An Exposition with Regard to the

Natural History of Animals, translated by Hugh Elliot from the 1st French edn. of 1809

- 33 -

with introductory essays by David L. Hull and Richard W. Burkhardt (Chicago: University

of Chicago Press).

Levitt, Barbara and March, James G. (1988) „Organizational Learning‟, Annual Review of

Sociology, 14, pp. 319-40.

Lewontin, Richard C. (1970) „The Units of Selection‟, Annual Review of Ecology and

Systematics, 1, pp. 1-18.

Maynard Smith, John and Szathmáry, Eors (1999) The Origins of Life: From the Birth of Life

to the Origin of Language (Oxford: Oxford University Press).

Mayr, Ernst (1960) „The Emergence of Evolutionary Novelties‟, in Tax, Sol (ed.) (1960)

Evolution After Darwin (I): The Evolution of Life (Chicago: University of Chicago Press),

pp. 349-80.

McKelvey, William (1982) Organizational Systematics: Taxonomy, Evolution, Classification

(Berkeley, CA: University of California Press).

Metcalfe, J. Stanley (1994) „Evolutionary Economics and Technology Policy‟, Economic

Journal, 104(4), July, pp. 931-44.

Nelson, Richard R. and Winter, Sidney G. (1982) An Evolutionary Theory of Economic

Change (Cambridge, MA: Harvard University Press).

Ouellette, Judith A. and Wood, Wendy (1998) „Habit and Intention in Everyday Life: The

Multiple Processes by which Past Behavior Predicts Future Behavior‟, Psychological

Bulletin, 124, pp. 54-74.

Peirce, Charles Sanders (1878) „How to Make Our Ideas Clear‟, Popular Science Monthly, 12,

January, pp. 286-302.

Polanyi, Michael (1967) The Tacit Dimension (London: Routledge and Kegan Paul).

Price, George R. (1995) „The Nature of Selection‟, Journal of Theoretical Biology, 175, pp.

389-96.

Reber, Arthur S. (1993) Implicit Learning and Tacit Knowledge: An Essay on the Cognitive

Unconscious (Oxford and New York: Oxford University Press).

- 34 -

Rigby, David L. and Essletzbichler, Jurgen (1997) „Evolution, Process Variety, and Regional

Trajectories of Technological Change in U.S. Manufacturing‟, Economic Geography,

73(3), pp. 269-284.

Robson, Arthur J. (1995) „The Evolution of Strategic Behaviour‟, Canadian Journal of

Economics, 28(1), pp. 17-41.

Rosenberg, Alexander (1992) „Neo-Classical Economics and Evolutionary Theory: Strange

Bedfellows?‟, PSA: Proceedings of the Biennial Meeting of the Philosophy of Science

Association1992, Volume One: Contributed Papers, pp. 174-183.

Simon, Herbert A. (1981) The Sciences of the Artificial, 2nd

edn. (Cambridge MA: MIT

Press).

Simon, Herbert A. (1990) „A Mechanism for Social Selection and Successful Altruism‟,

Science, 250, 21 December, pp. 1665-8.

Sperber, Dan (2000) „An Objection to the Memetic Approach to Culture‟, in Aunger, Robert

(ed.) (2000) Darwinizing Culture: The Status of Memetics as a Science (Oxford and New

York: Oxford University Press), pp. 162-73.

Stadler, Bärbel M. R., Stadler, Peter F., and Wagner, Günter P. (2001) „The Topology of the

Possible: Formal Spaces Underlying Patterns of Evolutionary Change‟, Journal of

Theoretical Biology, 213, pp. 241-274.

Steele, Edward J. (1979) Somatic Selection and Adaptive Evolution: On the Inheritance of

Acquired Characters (Toronto: Williams-Wallace International).

Steele, Edward J., Lindley, Robyn A., Blanden, Robert V. and Davies, Paul (1998) Lamarck’s

Signature: How Retrogenes are Changing Darwin’s Natural Selection Paradigm (New

York: Perseus).

Usher, John M. and Evans, Martin G. (1996) „Life and Death Along Gasoline Alley:

Darwinian and Lamarckian Processes in a Differentiating Population, Academy of

Management Journal, 39(5), October, pp. 1428-66.

Tomasello, Michael (2000) The Cultural Origins of Human Cognition (Cambridge, MA:

Harvard University Press).

van de Ven, Andrew H. and Poole, Marshall S. (1995) „Explaining Development and Change

in Organizations‟, The Academy of Management Review, 20(3), pp. 510-540.

- 35 -

Veblen, Thorstein B. (1914) The Instinct of Workmanship, and the State of the Industrial Arts

(New York: Macmillan).

Waddington, Conrad H. (1969) „The Theory of Evolution Today‟ in Koestler, Arthur and

Smythies, J. R. (eds) (1969) Beyond Reductionism: New Perspectives in the Life Sciences

(London: Hutchinson), pp. 357-74.

Waddington, Conrad H. (1976) „Evolution in the Sub-Human World‟, in Jantsch, Erich and

Waddington, Conrad H. (eds) (1976) Evolution and Consciousness: Human Systems in

Transition (Reading, MA: Addison-Wesley), pp. 11-15.

Weismann, August (1893) The Germ-Plasm: A Theory of Heredity, translated by W. Newton

Parker and Harriet R. Ronnfeldt (London and New York: Walter Scott and Scribner‟s).

Wilkins, John S. (2001) „The Appearance of Lamarckism in the Evolution of Culture‟, in

Laurent, John and Nightingale, John (eds) (2001) Darwinism and Evolutionary Economics

(Cheltenham: Edward Elgar), pp. 160-83.

Winter, Sidney G., Jr (1971) „Satisficing, Selection and the Innovating Remnant‟, Quarterly

Journal of Economics, 85(2), May, pp. 237-61.

Wood, Wendy, Quinn, Jeffrey M. and Kashy, D. (2002) „Habits in Everyday Life: Thought,

Emotion, and Action‟, Journal of Personality and Social Psychology, 83, pp. 1281-97.